When I think of
vibration analysis, I think less of bridges and more immediately of the catastrophic accident at Eschede, where 101 people died. After many years, the reasons have been identified, as well as the chain of tragic mistakes. I do not want to go into the details here. Nevertheless, I imagine permanent vibration analysis in all high-speed trains and the
axles of the trains and the
wheel flats, without cloud connection. Through machine learning, the Akida can perhaps filter the data far better than the current sensor technology can. So I'm thinking of a sensor specialist supplying train builders worldwide.
I add that are very important from an engineering point of view. As with the axles of the trains, the vibration analysis of
shafts for turbines is a very important thing. I think here first of all of shafts in
power plant turbines. And this in the internal network, think of cyber attacks. Then, of course, there are also turbines in
airplanes, for example. And then that brings me on to the analysis of airfoils. There are probably a thousand more topics for this use case. I hope a sensor specialist will soon be one of our customers or one of the customers of system suppliers who use our hardware or IP.
To this topic I add
earthquake detection and analysis. For example also on site at a volcano in a box. Machine learning could save the time for manual analysis. And maybe the detection is better, too. This brings to mind the severe earthquake in Italy. A funded earthquake early detection project of a very different kind is now underway there. It was found that animals became restless before the quakes. In the study, they are fitted with geo-transmitters to warn of a possible earthquake based on the data. With the vibration analysis of earthquakes and Akida, many lives could be saved worldwide.
I'm thinking about the wine testing video in the bar (funny video, I like that kind of thing) and the big incident just recently at Ferrero when masses of products were recalled and production was shut down. The only thing to blame was a single filter that was contaminated with salmonella. I imagine sensor technology that is trained with the Akida and then trains itself further through machine learning and thus automatically detects the smallest fluctuations, for example, when salmonella contaminates the product and slightly changes the composition.
Sensor technology and food industry and the analysis and detection of production fluctuations. When I think of the breath Covid test, a sensor in the food industry should also be able to detect bacteria or other impurities.
Akida in
FEM (Finite Element Method) to increase the efficiency of simulation technology through machine learning and AI. I have only read about this in theory so far. Data in industry is usually highly sensitive and must be protected in R&D departments. A link with a cloud is therefore a no go.
And Akida for automated
penetration testing or side channel analysis in IT security to test systems or devices for security and find vulnerabilities more efficiently with machine learning. And that without connecting to a cloud for similar reasons as above.
A bit of spinning.
Agile advertising. In some sci-fi movies and cities of the future, advertising is sometimes switched agilely. How would it be if large displays in the major cities of this world with the very good recognition of Akida display the advertising that is most interesting for the viewer. If a child stands in front of it, toys are shown, distinguishing between boys and girls. Or if a man stands in front of it, other advertising is shown than for women. Or again other advertising when a family, stroller, dog, skateboard or bicycle is recognized. And to top it off, Akida recognizes with Nvisio the facial expression whether the advertisement should be enthusiastic or better something else suggested.
Some may be familiar with the hassle of paying and waiting in line at
toll stations on the highway. Akidas recognition of the vehicle and automatic billing. And another topic:
speed traps for speeding and the recognition of the driver, self-sufficient in this unit.
Autonomous larger
cleaning vehicles or small robot depending on the application. At first, only on company premises or private property until people get used to it. Akida's detection of things on the ground correctly categorizes them and acts accordingly.
Adapted to Perhab's post: For me, the biggest strength of 5G is the minimal latency and automation from factory floors in real time from, Industry 4.0. I've been to a few symposiums on the topic. There was a lot of imagination in it. But with the recognition of Akida and I can imagine tremendous progress in
autonomous warehousing. No scanning with the scanner or manual declaring. One shot and the conveyor robot knows what to do, whether to sort into the warehouse or transport directly to the robot cell, where the Akida recognition takes over the parts. What used to be fantasy, I can imagine in reality with Akida.
Another bit of fun:
Akida at the poker table can certainly be an exciting experience for one or the other. Surely, all electronics will soon be banned at the tables as soon as someone has made his money with Akida. Besides, I imagine Akida IP in every non Apple and non Google phone anyway. Google immediately makes me think of Google Lens, but that's going too far.
I'm driving past a container terminal right now.
Fully automated autonomous loading and unloading at the world's
container terminals. Akida automatically detects the position and type of container and recognises the various means of transport, no matter what type.
The container ships lead me directly to the next topic.
Driving assistance systems for small
ships. Boris Herrmann would not have had a crash with Akida and would certainly have won the Vendée Globe. And I don't just mean a warning system, they have enough of those. But a pure emergency assistance system to avoid accidents and collisions. It was also reported how big the challenge is to recognise objects floating in the sea. This is where Akida could probably be very useful.
And tied to that, another theme for ships. The EQXX gives me the idea.
Supporting systems with machine learning
for ships to save fuel, depending on the situation, weather data, position, wind and waves.
Next topic, audio recognition. Sooner or later Akida will be in smartphones,) then my thought will be standard or obsolete. Until then, I imagine a
display device with Akida
for deaf people. They would only have to hold the device and read live on the display what is being said or announced at stations. As with the EQXX, machine learning also enables the recognition of dialects and filters out the clearest from the babble of speech, depending on the volume level or frequency, perhaps even selecting which layer via filters or settings.
I'll just keep going here until it becomes a possible use case book. Is there a character limit here I'm threatening to use this as my notebook.